Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
World Development Vol. xx, No. x, pp. xxx–xxx, 2010 Ó 2010 Elsevier Ltd. All rights reserved. 0305-750X/$ - see front matter www.elsevier.com/locate/worlddev doi:10.1016/j.worlddev.2009.06.011 The Role of Agriculture in African Development XINSHEN DIAO International Food Policy Research Institute, Washington, DC, USA PETER HAZELL University of London, UK and JAMES THURLOW * International Food Policy Research Institute, Washington, DC, USA Summary. — Widespread rural poverty in Africa and the success of Asia’s Green Revolution suggest that agriculture is a key sector for African development. However, in response to recent skepticism, this paper examines whether the conventional wisdom about agriculture’s contribution to the development process can still be applied to Africa today. We first outline the debate between the proponents of agriculture and its skeptics before presenting a series of case studies reflecting the heterogeneity of initial conditions facing low-income African countries. Drawing on economy-wide modeling, these case studies contrast the effectiveness of alternative growth strategies in reducing poverty. The findings indicate that while Africa does face many new challenges unlike those faced by Asian countries, there is little evidence to suggest that these countries can bypass a broad-based agricultural revolution to successfully launch their economic transformations. Ó 2010 Elsevier Ltd. All rights reserved. Key words — agriculture, rural development, poverty, Africa 1. INTRODUCTION 2. CONVENTIONAL WISDOM AND CURRENT DEBATE A majority of sub-Saharan Africa’s population live in rural areas where poverty and deprivation are the most severe. Since almost all rural households depend directly or indirectly on agriculture, and given the sector’s large contribution to the overall economy, it might seem obvious that agriculture should be a key sector in development. However, while agriculture-led growth has played an important role in reducing poverty and transforming the economies of many Asian countries, the strategy has not yet worked in Africa. Most African countries have failed to meet the requirements for a successful agricultural revolution, and productivity in African agriculture lags far behind the rest of the world. This has recently led to renewed debate within the international development community concerning the role of agriculture, particularly small farms, in African development. This paper considers whether the conventional wisdom about agriculture’s contribution to the development process can still be applied to Africa today. In Section 2 we briefly review the conventional wisdom and outline the current debate between the proponents of agriculture and its skeptics. Then, in Section 3, we present a series of case studies reflecting the heterogeneity of initial conditions facing low-income African countries. Drawing on economy-wide modeling, these case studies evaluate alternative growth strategies and their impact on poverty. The poverty-growth elasticities resulting from accelerating growth in different sectors are calculated in order to estimate the relative effectiveness of agriculture- and industry-led growth in reducing poverty. The paper concludes by summarizing the findings and drawing implications for the ongoing debate on the role of agriculture in African development. The perceived role of agriculture in growth and development has changed considerably over the last half-century. Building on the dual economy model, early theorists viewed economic development as a growth process requiring the reallocation of factors of production from a backward, low-productivity agricultural sector to a modern industrial sector with higher productivity and increasing returns (Lewis, 1954). As a traditional sector, agriculture was seen to contribute passively to development by providing labor and food to the industrialization process. However, this view was swept aside by the dynamism of the Green Revolution in Asia during the late 1960s and early 1970s. The possibility of transforming traditional agriculture into a modern sector demonstrated agriculture’s potential as a growth sector and its active role in initiating broader development (Adelman, 2001). 1 While the importance of linkages between agriculture and non-agriculture in driving the growth process had long been recognized (Hirschman, 1958; Johnston & Mellor, 1961), post-Green Revolution theorists emphasized the role of agriculture in rural development (Haggblade, Hammer, & Hazell, 1991; Haggblade, Hazell, & Brown, 1989; Hazell & Haggblade, 1991; Hazell & Roell, 1983). The positive impact of agricultural growth on rural development was found to be strongest in countries where small farms dominated agriculture (Rosegrant & Hazell, 2000). Therefore, given widespread rural poverty and small-scale farming in Africa, the “conventional wisdom” supports a strong role for agriculture in African development. * Final revision accepted: June 23, 2009. 1 Please cite this article in press as: Diao, X. et al. The Role of Agriculture in African Development, World Development (2010), doi:10.1016/j.worlddev.2009.06.011 2 WORLD DEVELOPMENT Today, there is a growing debate over whether traditional development theories still apply, and accordingly, whether agriculture can contribute significantly to growth in subSaharan Africa. Agriculture’s proponents suggest that, in many African countries, only the agricultural sector has sufficient scale and growth-linkages to significantly influence aggregate growth. Moreover, to significantly reduce poverty, it will be necessary to promote “shared growth” (Birdsall, Ross, & Sabot, 1995) and substantially raise incomes for a majority of Africa’s population. Achieving such growth will have to involve a large sector like agriculture, which accounts for one-third of gross domestic product (GDP) for the subcontinent as a whole, and an even larger share for two-thirds of African countries. Proponents also suggest that agriculture’s poor performance reflects inadequate investment and policies that are historically biased against the agricultural sector (Fan, Zhang, & Rao, 2004; Schiff & Valdez, 1992; Timmer, 2005). They emphasize the large gains from investing in rural infrastructure and agricultural technology, and the growth potential from catching up to the productivity levels of other developing countries. 2 Finally, proponents of agriculture suggest that there are few viable alternatives to agriculture, especially given many African countries’ small and poorly performing industrial sectors. During 1990–2004, African industry, including mining and mineral-based manufacturing, grew at 1.9% per year compared to 2.5% for agriculture (World Bank, 2006). 3 Africa also faces increasing competition from large emerging economies like China and India, which may undermine attempts to develop labor-intensive manufacturing sectors. By contrast, those skeptical of agriculture doubt whether agriculture can successfully generate sufficient growth in Africa today. This doubt reflects the poor performance of agriculture, weak institutions for rural development, and worsening agro-ecological conditions in many African countries (see, e.g., Collier, 2002; Ellis, 2005; Maxwell & Slater, 2003). The large size of the agricultural sector may be indicative of Africa’s failure to develop, especially since past experience predicts a significant decline in the importance of agriculture over time in successfully developing countries (Collier, 2002). Agriculture’s skeptics also suggest that while the sector’s strong growth-linkages proved very effective during Asia’s Green Revolution, they may be much weaker today in a more integrated global environment and with falling world food prices (Hart, 1998). Food prices are determined more by border prices than domestic supply when imports can enter freely, which reduces the need to invest in domestic agriculture to maintain urban food prices and real wages and hence industrial competitiveness. Under these emerging conditions, it is difficult for agriculture to generate economy-wide growth and facilitate the economic transformation predicted by theory or witnessed in the past successes of other developing countries. Agriculture’s skeptics therefore tend to be more optimistic of African industry, suggesting that mining and manufacturing may offer viable alternative sources of growth. Despite contrasting opinions on the relative importance of agriculture and industry in generating overall growth, there should presumably be less contention surrounding the role of agriculture in poverty reduction, especially given the importance of agricultural incomes for Africa’s poorest populations. However, even among agriculture’s proponents, there are conflicting opinions over what should be the focus of an agricultural development strategy for low-income Africa (Dorward, Kydd, Morrison, & Urey, 2004). Some argue that the best opportunities for African farmers lie in high-value commodities and, given weak domestic demand in Africa, that produc- tion should focus on export markets. Small-scale farms are seen as unviable due to international competition and the growing complexity of supply-chains for both domestic and foreign markets (Reardon, Timmer, Barrett, & Berdeguè, 2003). Rural populations should therefore focus on diversifying incomes away from agriculture (Ashley & Maxwell, 2001) and migrating to urban areas (Ellis & Harris, 2004). By contrast, others argue that rural income diversification has been a reality in Africa for decades (Barrett & Reardon, 2000; Reardon et al., 2003) and has yet to generate significant income growth. Furthermore, income diversification is not an unequivocally positive phenomenon, especially if diversification is driven by stagnant agricultural growth (Haggblade, Hazell, & Reardon, 2002) or if migration is due to growth in low-productivity urban activities (Lipton, 2004). Furthermore, while opportunities exist for improving traditional exports through better quality and niche markets, and while nontraditional exports are growing fast, the contribution of such growth to the overall economic growth is modest (Diao & Dorosh, 2007). The greatest market potential for most African farmers lies in domestic and regional markets for staples/food crops (Diao & Hazell, 2004; Rosegrant, Paisner, Meijer, & Witcover, 2001). With increasing commercialization and urbanization, future demand for these commodities will translate into market transactions and not just on-farm consumption. Therefore, the current debate is twofold, focusing first on the potential role of agriculture and industry in fostering African development and second, within agriculture, on the ability of the food and export sectors to generate pro-poor growth. This paper contributes to this debate by examining how accelerating growth in the sectors emphasized by agriculture’s proponents and skeptics can influence a country’s ability to significantly reduce poverty. Greater resolution in the debate is possible if it is recognized that country context matters and agriculture have different roles to play in different circumstances. For example, the agricultural sector as a whole and food staples in particular will have more important roles to play during the early stages of economic development when they dominate national income and employment. However, these sectors will undoubtedly diminish as development proceeds and agriculture becomes a relatively small sector. Agriculture is also likely to play a bigger role in countries with good agro-ecological conditions and limited prospects for export earnings from minerals or other industrial goods. While the role of agriculture may be more important in countries dominated by small farms, countries with dynamic and growing national economies and rising per capita incomes may offer farmers greater opportunities to diversify into higher value products and non-farm sources of income. Conversely, in poorer and slower growing economies, opportunities for income diversification and exit strategies are more limited. Given the importance of initial conditions, and in order to bring greater contextual specificity to the debate, in the next section we examine the relationship between agricultural growth and poverty reduction in six low-income sub-Saharan African countries. 3. AGRICULTURAL GROWTH AND POVERTY REDUCTION: SELECTED COUNTRY CASE STUDIES (a) Selection of case study countries Six case study countries – Ethiopia, Ghana, Kenya, Rwanda, Uganda, and Zambia – were selected to reflect a range Please cite this article in press as: Diao, X. et al. The Role of Agriculture in African Development, World Development (2010), doi:10.1016/j.worlddev.2009.06.011 THE ROLE OF AGRICULTURE IN AFRICAN DEVELOPMENT 3 Table 1. Comparative indicators across selected case study countries. Source: World development indicators (World Bank, 2006); national poverty rates from various sources (Diao et al., 2007b) Classification GDP per capita, 2004 ($US) Share of GDP, 2004 (%) Agriculture Industry Manufacturing GDP growth rate, 1990–2004 (%) Agriculture Industry Rural population, 2004 (%) Poverty headcount (%) Dollar-a-day, 1999 National poverty line (Survey year) Ethiopia Ghana Kenya Rwanda Uganda Zambia Inland 114.4 Coastal 409.4 Coastal 480.7 Inland 207.7 Inland 245.2 Mineral 470.6 46.9 9.5 6.1 3.5 2.1 2.1 84.1 37.9 24.7 8.5 4.5 3.8 3.2 54.2 26.8 17.2 11.1 2.2 1.8 1.6 59.5 40.5 21.5 10.0 1.6 3.4 -1.3 79.9 32.2 21.2 9.2 6.5 3.8 9.9 87.7 20.9 26.9 12.0 1.7 2.6 0.0 63.8 85.2 44.2 (1999/00) 44.8 39.5 (1998/99) 23.9 47.1 (1997) 58.9 60.3 (1999/01) 40.8 35.0 (1999/00) 79.3 75.4 (1998) of initial conditions in Africa (cf. Table 1). Both Ethiopia and Rwanda are land-locked countries with some of the highest population densities in Africa. In both countries, agriculture accounts for a substantial share of gross domestic product (GDP) and approximately four-fifths of the population live in rural areas. Furthermore, agricultural production is dominated by food crops or traditional livestock. Moreover, Rwanda’s high altitude means that the country has less favorable agricultural conditions relative to other African countries (Diao, Hazell, Resnick, & Thurlow, 2007b). By contrast, Ghana and Kenya are coastal countries with more favorable agricultural conditions. Both countries have significant agricultural export sectors (e.g., cocoa in Ghana and tea and cut flowers in Kenya). Together with Uganda, they have relatively large and growing industrial sectors with more established agro-processing and other manufacturing. In Zambia, agriculture’s share of GDP is smaller than industry’s, reflecting the country’s considerable mineral resources. Although agricultural exports, such as cotton, have grown rapidly, they remain small compared to more traditional food crops. Finally, Uganda is also a land-locked country with favorable agricultural conditions. Together with Ghana, the two countries are among the few African countries that have experienced high and stable GDP and agricultural growth over a sustained period. (b) Modeling growth and poverty reduction The country studies use economy-wide simulation models to examine the relative contribution of agriculture to poverty reduction and growth. Two types of models are used: economy-wide, multimarket (EMM) models for Ethiopia, Ghana, and Rwanda and computable general equilibrium (CGE) models for Kenya, Uganda, and Zambia. 4 As mentioned earlier, an important factor determining the contribution of agriculture to economic growth is the linkages between agriculture and the rest of the economy. Agriculture is found to have strong growth-linkages in many countries. Economy-wide models are therefore chosen for the case studies in order to capture these linkages in both consumption and production processes. To achieve this, the models are calibrated to highly disaggregated datasets that include detailed agricultural and non-agricultural productive activities. Each country’s economy is further disaggregated across sub-national regions. This is important because the growth-poverty linkages may be different in remote rural regions with high poverty, compared to urbanized regions with larger industrial concentrations and urban employment. Another focal area of the debate relates to the role of agriculture in an integrated world environment. This debate concerns that cheap food imports may undermine agriculture’s growth-linkages within countries and reduce the need for domestic agricultural production. Furthermore, the debate concerns the larger export market opportunities of high-value crops and their linkages with overall growth. To help assess these issues, the models capture both import competition and export opportunities by allowing producers and consumers to shift between domestic and foreign markets depending on changes in relative prices. A third area in the debate concerns diversification in rural household livelihoods. The models capture detailed income and expenditure patterns across sub-national regions, rural and urban areas, and different types of households. Each country’s Living Standards Monitoring Survey (or its equivalent) is used to define representative households, each of which is an aggregation of a group of households in the surveys. In order to retain enough information on household heterogeneity, the economy-wide models are linked to survey-based microsimulation modules. Endogenous changes in either consumption expenditure or sources of income for each representative household in the models are mapped onto the commodity expenditure or income source of the corresponding households in the surveys. Such linkages allow us to estimate poverty-growth elasticities at both the national and sub-national levels. In summary, the models capture (i) disaggregated growth across regions and sectors; (ii) employment effects through factor markets and price effects through commodity markets within countries and through foreign trade; and (iii) household-level income and poverty effects according to either income sources or expenditure patterns. (c) Agricultural growth is more pro-poor than industrial growth A baseline scenario is first simulated in which the six countries are assumed to continue growing according to current trends until 2015. These trends include not only the level of aggregate economic growth but also its sectoral composition. It is widely held that most African countries are unlikely to meet with the first Millennium Development Goal (MDG) of halving poverty by 2015 unless their growth performance improves dramatically. The models’ results confirm this, with the exception of Ghana and Uganda, where halving poverty is Please cite this article in press as: Diao, X. et al. The Role of Agriculture in African Development, World Development (2010), doi:10.1016/j.worlddev.2009.06.011 4 WORLD DEVELOPMENT achievable under the two countries’ current growth trends. 5 Taking Ethiopia as an example, the baseline scenario shows that if the current level and composition of growth are maintained, then the poverty headcount rate is likely to remain unchanged at around 44.3% by 2015 (cf. Table 2). Ethiopia must therefore not only accelerate the level of growth but also find ways in which to enhance its “pro-poorness.” In other words, it is also necessary to identify the composition of growth that is most effective at reducing poverty and that raises the poverty-growth elasticity. In the context of the current debate, this begs an examination of the relative importance of agriculture and industry in helping Africa achieve its development objective of significantly reducing poverty. To this end, the models are used to generate two scenarios in which agricultural and industrial growth are accelerated separately and the effectiveness of this additional growth in reduc- ing poverty is compared. To make the results comparable, poverty-growth elasticities are calculated for each scenario in the six countries. The results show that the poverty-growth elasticity is consistently larger when growth is driven more by agricultural growth than non-agricultural growth (cf. Table 2). Our results are consistent with those of Christiaensen, Demery, and Kühl (2006), who use a pooled sample of subSaharan African countries and econometrically show that agricultural growth has a greater poverty reduction effect than non-agricultural growth. As an example, in our model a 1% annual increase in Ethiopia’s per capita GDP driven by agricultural growth leads to a 1.7% reduction in the country’s poverty headcount rate per year. By contrast, a similar increase in per capita GDP driven by non-agriculture in the country leads to only a 0.7% reduction in the poverty rate. Over time, these deviations in the poverty-growth elasticity can translate into Table 2. Comparison of agricultural and non-agricultural growth scenarios. Source: Authors’ simulations and calculations Baseline scenario Agricultural-led scenario Non-agricultural-led scenario Ethiopia (2003–15) Annual per capita GDP growth rate (%) Annual GDP growth rate (%) Agriculture Non-agriculture Poverty headcount by 2015 (%) Poverty-growth elasticity 0.5 3.1 2.5 3.7 44.3 – 2.4 5.0 5.0 5.0 26.5 1.66 2.4 5.0 2.7 7.0 37.3 0.73 Ghana (2003–15) Annual per capita GDP growth rate (%) Annual GDP growth rate (%) Agriculture Non-agriculture Poverty headcount by 2015 (%) Poverty-growth elasticity 2.2 4.7 4.6 4.8 23.7 1.49 3.1 5.7 7.0 4.8 17.3 1.78 3.1 5.7 4.6 6.2 21.5 1.33 Kenya (2003–15) Annual per capita GDP growth rate (%) Annual GDP growth rate (%) Agriculture Non-agriculture Poverty headcount by 2015 (%) Poverty-growth elasticity 1.1 3.0 3.4 2.9 46.2 0.67 2.1 4.0 6.1 3.4 36.0 1.25 2.1 4.0 3.6 4.2 44.1 0.57 Rwanda (2003–15) Annual per capita GDP growth rate (%) Annual GDP growth rate (%) Agriculture Non-agriculture Poverty headcount by 2015 (%) Poverty-growth elasticity 0.7 3.4 3.3 3.4 55.5 1.09 3.2 6.0 7.9 3.5 34.6 1.41 3.2 6.0 3.5 8.1 43.3 0.84 Uganda (1999–2015) Annual per capita GDP growth rate (%) Annual GDP growth rate (%) Agriculture Non-agriculture Poverty headcount by 2015 (%) Poverty-growth elasticity 1.6 5.2 5.1 5.3 27.8 0.98 2.8 6.4 7.6 5.2 17.6 1.58 2.8 6.4 5.3 7.4 21.7 1.10 Zambia (2001–15) Annual per capita GDP growth rate (%) Annual GDP growth rate (%) Agriculture Non-agriculture Poverty headcount by 2015 (%) Poverty-growth elasticity 2.0 4.0 4.6 3.8 68.3 0.35 3.0 5.0 7.7 4.0 58.9 0.58 3.0 5.0 4.5 5.1 64.4 0.38 Notes: Poverty rates are measured according to nationally defined poverty lines; the non-agricultural simulations for Kenya and Zambia involved accelerating growth in only the industrial sectors. Please cite this article in press as: Diao, X. et al. The Role of Agriculture in African Development, World Development (2010), doi:10.1016/j.worlddev.2009.06.011 THE ROLE OF AGRICULTURE IN AFRICAN DEVELOPMENT significantly different poverty outcomes. With similar GDP growth, the poverty headcount in Ethiopia falls to 26.5% under the agriculture-led growth scenario, compared with 37.3% under the non-agriculture-led growth scenario. Given its larger impact on poverty, agriculture-led growth in Ethiopia lifts an additional 9.6 million people out of poverty compared to nonagriculture-led growth, despite the fact that overall GDP grows at the same rate under the two scenarios. These findings are consistent across the six countries studied. Given a similar GDP growth rate, the calculated poverty-growth elasticities are always higher under the agriculture-led scenario. However, the magnitudes of these differences vary across countries. Why does agricultural growth do so much better in benefiting the poor? Part of the reason arises because growth affects individuals differently due to heterogeneity across regions and households. Given different income sources, consumption patterns, and locations within a country, changes in total income and consumption at the individual household level differ considerably from average changes at the national level (i.e., per capita GDP or total consumption). These differences are country-specific due to variations in initial economic structures and income distributions. For example, in some countries agriculture generates a large share of national GDP and many households live in rural regions dominated by agriculture. For these households, participation in agricultural activities is often the major source of income, and hence they are likely to benefit more from agriculture-led growth than non-agricultural-led growth. Households with greater opportunities to work in the urban sector or to take advantage of near-by city markets for higher value agricultural products may concentrate in periurban areas and be better positioned to benefit from non-agriculture or export agriculture. Since such households are usually less poor than remoter households, economic growth driven by non-agriculture may have a smaller impact on poverty reduction. For example, according to the Rwandan national household survey conducted in 2000–01, agriculture accounts for half of the average household income at the national level, but three-quarters of income for the average poor household. The importance of agricultural incomes is even greater in poorer regions of the country. Under such circumstances, agricultural growth is expected to be more pro-poor than non-agricultural growth since agriculture is a more important income source for the poor. Another reason is that agricultural growth can also benefit urban, landless, or “net food buyer” rural households if rising agricultural productivity lowers food prices. While globalization and trade liberalization allow cheaper imported foods to enter more easily into metropolitan markets, most households in rural areas and small towns still rely on domestically supplied food staples, primarily due to high transaction and transport costs. 6 The models account for price differentials among food staples, either within a country or between imported and domestic goods, such that international prices cannot fully transform into domestic prices. Accordingly, when domestic food supply increases in response to higher agricultural productivity, the result is a decline in food staple prices. While this benefits consumers in general, its impact on poor households is particularly strong since food purchases are the major items in their consumption basket. For example, Ethiopia’s 1999–2000 national household survey showed that poor urban households spend, on average, more than 50% of their total income on staple foods, which is higher than the corresponding 30% for all urban households. Therefore, initial conditions are the primary factors determining the size of the poverty-growth elasticity in each country. For example, with perfect labor markets, rural households 5 can benefit more from non-agricultural growth by migrating to urban areas and participating in urban-based industrial activities. However, many economic, institutional, social, and cultural barriers constrain such labor movement. Because of this, an imperfect labor market is assumed in our models, and hence, the poor rural households have limited opportunities to participate in urban-based growth. The large gap between the poverty-growth elasticities in the two scenarios indicates the relative importance of agricultural growth, especially for poorer rural households (cf. Table 2). This is true even for a country like Zambia, in which agriculture generates only a small share of GDP. Zambia’s economic structure partly reflects the country’s long-standing dependence on copper production and exports, which has fostered a dual economy biased in favor of urban-based industrialization. Copper mining is a capital-intensive enclave industry with weak linkages to the rest of economy, especially the rural economy. History has shown that growth driven by this sector does not provide the magnitude of poverty reduction needed in this impoverished country. The model simulations for Zambia support this and show that growth in the non-agricultural sector, led by the mining industrial sectors, is less effective at reducing poverty than agriculture-led growth. For instance, with similar overall economic growth, growth led by non-agriculture would reduce poverty in Zambia to 64.4% by 2015, compared with 58.9% by the same year under an agriculture-led growth scenario. While one of the key arguments presented by the proponents of agriculture is that this sector has sufficient scale to generate significant economy-wide growth, our model results suggest that this is not the only reason to support agriculture’s key role in development. The two scenarios mentioned above were designed as “scale-neutral” by targeting the same total GDP growth rate in each individual country. The model results show that agriculture has a greater ability to generate employment and incomes among the poor population and provide broad-based growth that is shared by both poor and non-poor households. Focusing on the aggregate growth alone may not significantly reduce poverty if growth generates distributional changes that isolate the poor from the growth process. For this reason, the model results support an important role for agriculture in African development. (d) Broad-based agricultural growth is more pro-poor than export-led growth The second part of the current debate concerns the relative importance of food crops and agricultural exports in agricultural growth and poverty reduction. Agricultural exports have grown rapidly in recent years in many African countries. However, while high-value agriculture may have greater growth potential, its contribution to economy-wide growth may be insufficient over the foreseeable future due to its small initial base in most African countries. Moreover, as mentioned above, growth in high-value export crops may only reach a small group of farmers with better urban and export market access and often has less impact on the food costs of the poor. For instance, food crops compose 65% of agriculture in Ethiopia, and a majority of the poor depend heavily on cultivating these crops. This is equally true in Rwanda, where the share of staple crops and livestock in total agricultural output is as high as 90%. The share of staples and livestock is lower in the other case study countries, but is still 70% of total agricultural output in Ghana, 54% in Uganda and Kenya, and 65% in Zambia. Therefore, the extent to which agricultural sub-sectors are able to contribute significantly to aggregate growth and Please cite this article in press as: Diao, X. et al. The Role of Agriculture in African Development, World Development (2010), doi:10.1016/j.worlddev.2009.06.011 6 WORLD DEVELOPMENT poverty reduction varies considerably. In line with the current debate, this section examines two broad sub-sectors in terms of their ability to reduce poverty: staple food, including staple crops and livestock; and export crops, including both traditional and nontraditional crops. Assuming similar growth rates at the sub-sector level, greater economy-wide growth will obviously be generated by the larger sub-sector, in turn, producing a (generally) larger effect on poverty. On the other hand, small sub-sectors, such as nontraditional export crops, may have greater capacity to grow rapidly. Therefore, in determining whether a sub-sector can ultimately “drive” growth, both economy-wide linkages and market potential (determined by supply and demand factors) should be considered. In order to ensure that the two simulations are comparable despite having different initial contributions to GDP, it is necessary to accelerate growth in each subsector until a similar aggregate growth rate is achieved. Taking Zambia as an example, in order for growth led by export crops to generate an additional 1% annual growth in national GDP (i.e., from 4% to 5%), these crops would have to grow at 23% per year because this sub-sector is initially very small and has relatively weak growth spillovers to other sectors (cf. Table 3). By contrast, the staples sector is substantially larger and links with other sectors through domestic markets. Thus, this subsector does not have to grow as rapidly to produce the same additional 1% annual growth in GDP. Similarly, to achieve 5% annual growth in agricultural GDP in Ethiopia, the required growth rate for staple crops and livestock is 5% if additional agricultural growth is driven by the staples sub-sector. However, an annual growth of export crops of 18% is required for those crops to drive the same agricultural growth rate. This high growth requirement for export crops is found for each of the case study countries. Although export sectors undoubtedly have considerable growth potential, it is unlikely that such high growth rates in any sector are feasible over a sustained period. Table 3. Comparison of staples and exportable agricultural growth scenarios. Source: Authors’ simulations and calculations Baseline scenario Staple-crops-led scenario Export-crops-led scenario Ethiopia (2003–15) Annual per capita GDP growth rate (%) Annual GDP growth rate (%) Agriculture Staple crops Export crops Poverty headcount by 2015 (%) Poverty-growth elasticity 0.5 3.1 2.5 2.0 4.0 44.3 – 2.4 5.0 5.0 5.0 4.4 27.2 1.80 2.4 5.0 5.0 1.9 18.0 31.6 1.40 Ghana (2003–15) Annual per capita GDP growth rate (%) Annual GDP growth rate (%) Agriculture Staple crops Export crops Poverty headcount by 2015 (%) Poverty-growth elasticity 2.2 4.7 4.6 4.6 4.1 23.7 1.50 3.4 6.0 7.7 8.5 3.4 14.0 2.10 3.4 6.0 7.7 3.7 18.4 22.9 1.10 Kenya (2003–15) Annual per capita GDP growth rate (%) Annual GDP growth rate (%) Agriculture Staple crops Export crops Poverty headcount by 2015 (%) Poverty-growth elasticity 1.1 3.0 3.4 1.7 5.5 46.2 0.67 2.1 4.0 5.4 7.8 1.7 36.2 1.19 2.1 4.0 6.4 1.3 10.1 36.5 1.19 Uganda (1999–2015) Annual per capita GDP growth rate (%) Annual GDP growth rate (%) Agriculture Staple crops Export crops Poverty headcount by 2015 (%) Poverty-growth elasticity 5.2 5.1 5.1 4.4 27.8 0.98 6.4 7.7 9.0 1.6 18.6 1.40 6.4 7.9 5.0 19.7 19.0 1.39 Zambia (2001–15) Annual per capita GDP growth rate (%) Annual GDP growth rate (%) Agriculture Staple crops Export crops Poverty headcount by 2015 (%) Poverty-growth elasticity 4.0 4.6 4.1 10.2 68.3 0.35 5.0 7.8 7.9 6.9 59.2 0.57 5.0 7.1 4.0 22.8 62.0 0.47 Notes: Poverty rates are measured according to nationally defined poverty lines; since the agricultural export sector is very small in Rwanda, it is not included in these simulations. Please cite this article in press as: Diao, X. et al. The Role of Agriculture in African Development, World Development (2010), doi:10.1016/j.worlddev.2009.06.011 THE ROLE OF AGRICULTURE IN AFRICAN DEVELOPMENT Growth in the staple sector has a larger impact on agricultural and economic growth because it is a larger sector than export agriculture and has stronger growth-linkages. However, staples growth also leads to strongly pro-poor outcomes because of this sector’s broad base. The results show that even if extremely high growth in export crops is possible, it leads to much smaller poverty-growth elasticities. For example, if the same 5% agricultural GDP growth rate in Ethiopia is driven by the staples sector, then the national poverty rate is likely to fall to 27% by 2015. This is 4.4 percentage points lower than the poverty rate expected under the agricultural-export-led scenario, despite these two scenarios targeting the same 5% agricultural growth rate. This greatly improves poverty outcomes, lifting an additional four million people above the poverty line by 2015. There is concern over market constraints for food staples since these commodities are often income inelastic. Some skeptics suggest that increasing food supply will hurt farmers because prices may fall sharply. However, even though rapid declines in staples prices have been observed in some African countries during good harvest years, there is a rising trend in imported staple crops, such as maize and rice, indicating potential markets for increased domestic supply of these commodities. 7 The models capture these observed import trends in the baseline scenarios, since an expanding population and rapid urbanization generate domestic demand for staple foods that cannot be met solely by domestic supply. Therefore, taking market constraints into account, the model results still indicate that staple crops and livestock have sufficient scale to foster economy-wide growth and significantly reduce poverty in many African countries. By contrast, in order for agricultural exports to contribute significantly to aggregate growth, the sub-sector would have to achieve and sustain very high growth rates. While these export sectors may well have greater growth potential, it is unrealistic to be too optimistic about the potential for agricultural exports, especially given the past volatility in world prices for agricultural commodities. Furthermore, even if export agriculture could sustain such high growth, the sector is less effective at reducing poverty than are staple crops and livestock. Therefore, while the results do not suggest that agricultural exports hamper poverty reduction, they do emphasize the importance of investing in food staples and in designing rural development strategies around broad-based agricultural growth. (e) Is broad-based agricultural growth achievable? The results from the country case studies indicate that, at least in the short term, both agricultural and staple-led growth will have to accelerate if there is to be significant growth and poverty reduction in Africa. However, the past poor performance of the staple sector in many African countries raises doubts over whether this growth is attainable. For instance, while Africa’s staple sector may be critical in reducing poverty, past growth in this sector has typically arisen from area expansion. Until the 1970s, many African countries were considered self-sufficient in food crop production and there seemed little need to pay attention to the food sector or change established methods of production. However, this situation has changed dramatically over the last three decades. Expansion of arable land has stagnated in recent years, indicating that land frontiers may have been reached. The result is mounting population pressure and declining farm sizes. The land constraint is particularly serious in countries like Ethiopia and Rwanda, where land distribution and farm sizes are worse than in many Asian countries at the time of the Green Revolution. However, 7 while land expansion has dominated past growth, there is an extensive literature identifying the potential for intensifying food crop production in Africa. For example, Djurfeldt, Holmen, Jirstrom, and Larsson (2005) conducted surveys of 3,000 small-scale farmers in eight African countries (including Ethiopia, Ghana, Kenya, Uganda, and Zambia) and found evidence where farms or regions have achieved crop yields far above the national average. Other studies also support the finding that intensification has occurred in particular African regions and among certain categories of farmers (Gabre-Madhin & Haggblade, 2001; Haggblade et al., 2002; Turner, Hyden, & Kates, 1993; Wiggins, 1995, 2000). Therefore, while Africa has not achieved its Green Revolution, there are numerous examples where the intensification of food crop production has been successful. Transforming individual success stories into broader agricultural development remains a challenge. Innovations in science and technology are both a precondition and a part of agricultural transformation. African farmers need technologies applicable to their diversified agro-ecological conditions in order to counter erratic rainfall and declining soil fertility. However, while technologies suitable for African agriculture are crucial in shifting from land expansion to intensification, Africa cannot rely on the package of technologies in agricultural production only. Agricultural intensification also involves the development of supply chains around small-scale farmers, such as the presence of input markets, seasonal finance, and marketing systems to increase farm production and deliver goods to consumers at competitive prices (Poulton, Kydd, & Dorward, 2006). However, a lack of profitable opportunities will deter private sector engagement. Accordingly, public investments are needed in agricultural research and in market and institutional development in order to reduce costs and mitigate risks. Numerous studies confirm that increased investment in agricultural research is required to revitalize agricultural development (Alston, Norton, & Pardey, 1998; Fan et al., 2004). Rural infrastructure is also necessary to increase consumer demand and farmers’ access to input and output markets, to stimulate the rural non-farm economy and rural towns, and to more fully integrate the poorest regions into their countries’ economies. Spencer (1994) estimates that the rural road density in Africa is well below Asia’s at the time of the Green Revolution. However, while substantial investment is needed throughout much of Africa, Fan et al. (2004) have shown in Uganda that rural investments do not have to be excessive in order to have sizeable impacts. Therefore, there is evidence both suggesting that broad-based agricultural growth is possible in Africa and identifying key interventions to achieve this growth. 4. CONCLUSION This paper has responded to the current debate concerning the role of agriculture in African development by examining the impact of alternative growth paths on overall growth and poverty reduction in six low-income African countries. The findings suggest that the structure of growth is important in determining not only the total growth but also the size of a country’s poverty-growth elasticity. Although some African countries have potential sources of growth outside of agriculture, growth in the industrial sectors is, at least in the short- to medium-term, unlikely to be substantial enough in many countries. Moreover, non-agricultural growth is generally less effective at reducing poverty as agricultural growth. While the large size of the agricultural sector predetermines its potential Please cite this article in press as: Diao, X. et al. The Role of Agriculture in African Development, World Development (2010), doi:10.1016/j.worlddev.2009.06.011 8 WORLD DEVELOPMENT influence on aggregate growth, we have looked beyond the size of the sector and compared the nature of different sectors’ growth-poverty linkages. Our analysis shows that agricultural growth is more pro-poor than industrial growth, primarily because it allows for greater participation of the poor in the growth process. While the industrial sector also has linkages to the rest of the economy, these linkages are weaker and create fewer employment opportunities for poor workers. This is similarly true for agricultural exports, which typically benefit peri-urban areas and do not necessarily benefit the poor in more remote rural areas. The paper has also reviewed the literature and finds numerous cases where there is both potential for greater intensification in staples production and evidence to support key interventions. The findings therefore indicate that, while Africa does face many new challenges unlike those faced by Asian countries, there is little evidence to suggest that these countries can bypass a broad-based agricultural revolution to successfully launch their economic transformations. NOTES 1. A recent article by Blench and Verner (2006) examined agricultural and non-agricultural growth over three decades in Cote d’Ivoire, Ghana, and Zimbabwe and found a large degree of interdependence in long-term sectoral growth, implying that the sectors grow together and that agriculture is an engine of growth in the long term. 2. The average cereal yield in Africa currently is about a quarter of that in South Asia, despite these two regions having similar yields in the early 1960s (FAO, 2002). 3. Based on total unadjusted dollars summed across countries in 2000 prices. Due to missing data, this measure includes 26 out of 38 lowincome countries that accounted for four of the agricultural and industrial GDP in 1990. 5. See Al-Hassan and Diao (2007), Diao, Fan, Kanyarukiga, and Yu (2007a), Diao and Pratt (2007), Thurlow, Kiringai, and Gautam (2007) and Thurlow and Wobst (2006) for a description of the baseline scenarios for Ethiopia, Ghana, Kenya, Rwanda, and Zambia. 6. This excludes food aid, which often enters countries alongside foreign food supply. 7. The models do not simulate production shocks, such as a price collapses after a sudden increase in crop production due to good weather. Such shocks should be handled through appropriate price stabilization policies rather than constraints on agricultural production. 4. The CGE approach is preferable but constrained by data availability. See Diao et al. (2007b) for a detailed description of the models. REFERENCES Adelman, I. (2001). Fallacies in development theory and their implications for policy. In G. M. Meier, & J. E. Stiglitz (Eds.), Frontiers of development economics: The future in perspective (pp. 103–135). New York: World Bank. Al-Hassan, R. M., & Diao, X. (2007). Reducing regional disparities in growth and poverty reduction in Ghana: Policy options and public investment implications. IFPRI discussion paper no. 693. Washington, DC: IFPRI. Alston, J. M., Norton, G. W., & Pardey, P. G. (1998). Science under scarcity. New York and Wallingford: CAB International. Ashley, C., & Maxwell, S. (2001). Rethinking rural development. Development Policy Review, 19(4), 395–425. Barrett, C., & Reardon, T. (2000). Asset, activity, and income diversification among African agriculturalists: Some practical issues. Discussion paper. Ithaca, New York: Department of Agricultural, Resource and Managerial Economics, Cornell University. Birdsall, N., Ross, D., & Sabot, R. (1995). Inequality and growth reconsidered: Lessons from East Asia. World Bank Economic Review, 9(3), 477–508. Blench, N., & Verner, D. (2006). Shared sectoral growth versus the dual economy model: Evidence from Cote d’Ivoire, Ghana, and Zimbabwe. Africa Development Review, 18(3), 283–308. Christiaensen, L., Demery, L., & Kühl, J. (2006). The role of agriculture in poverty reduction: An empirical perspective. Policy research working paper series 4013. World Bank. Collier, P. (2002). Primary commodity dependence and Africa’s future. In B. Pleskovic (Ed.), Annual World Bank conference on development economics proceedings (pp. 29–30). Washington, DC: World Bank. Diao, X., & Dorosh, P. (2007). Demand constraints on agricultural growth in East and Southern Africa: A general equilibrium analysis. Development Policy Review, 25(3), 275–292. Diao, X., Fan, S., Kanyarukiga, S., & Yu, B. (2007a). Agricultural growth and investment options for poverty reduction in Rwanda. IFPRI discussion paper no. 689. Washington, DC: IFPRI. Diao, X., & Hazell, P. (2004). Exploring market opportunities for African smallholders. In 2020 Africa conference brief 6. Washington, DC: IFPRI. Diao, X., Hazell, P., Resnick, D., & Thurlow, J. (2007b). The role of agriculture in development: Implications for sub-Saharan Africa. Research report 153. Washington, DC: IFPRI. Diao, X., & Pratt, A. N. (2007). Growth options and poverty reduction in Ethiopia – An economy-wide model analysis. Food Policy, 32(2), 205–228. Djurfeldt, G., Holmen, H., Jirstrom, M., & Larsson, R. (2005). The African food crisis: Lessons from the Asian Green Revolution. Wallingford: CABI Publishing. Dorward, A., Kydd, J., Morrison, J., & Urey, I. (2004). A policy agenda for pro-poor agricultural growth. World Development, 32(1), 73–89. Ellis, F. (2005). Small farm, livelihood diversification, and rural–urban transitions: Strategic issues in sub-Saharan Africa. In The future of small farms, research workshop proceedings, Wye, UK, June 26–29, 2005 (pp. 135–149). Ellis, F., & Harris, N. (2004). New thinking about urban and rural development. Keynote paper for DFID sustainable development retreat. Fan, S., Zhang, X., & Rao, N. (2004). Public expenditure, growth and poverty reduction in rural Uganda. DSG discussion paper no. 4. Washington, DC: IFPRI. FAO (2002). FAOSTAT. United Nations, Rome: Food and Agriculture Organization. Gabre-Madhin, E. Z., & Haggblade, S. (2001). Success in African agriculture: Results of an expert survey. MSSD discussion paper 53. Washington, DC: IFPRI. Please cite this article in press as: Diao, X. et al. The Role of Agriculture in African Development, World Development (2010), doi:10.1016/j.worlddev.2009.06.011 THE ROLE OF AGRICULTURE IN AFRICAN DEVELOPMENT Haggblade, S., Hazell, P., & Brown, J. (1989). Farm-nonfarm linkages in rural sub-Saharan Africa. World Development, 17(8), 1173–1202. Haggblade, S., Hammer, J., & Hazell, P. (1991). Modeling agricultural growth multipliers. American Journal of Agricultural Economics, 73(2), 361–374. Haggblade, S., Hazell, P., & Reardon, T. (2002). Strategies for stimulating growth of the rural nonfarm economy in developing countries. EPDT discussion paper no. 92. Washington, DC: IFPRI. Hart, G. (1998). Regional linkages in the era of liberalization: A critique of the new agrarian optimism. Development and Change, 29(1), 27–54. Hazell, P., & Haggblade, S. (1991). Rural–urban growth linkages in India. Indian Journal of Agricultural Economics, 46(4), 512–529. Hazell, P., & Roell, A. (1983). Rural growth linkages: Household expenditure patterns in Malaysia and Nigeria. Research report. Washington, DC: IFPRI. Hirschman, A. O. (1958). The strategy of economic development. New Haven, Connecticut: Yale University Press. Johnston, D. G., & Mellor, J. W. (1961). The role of agriculture in economic development. American Economic Review, 51(4), 566–593. Lewis, W. A. (1954). Economic development with unlimited supplies of labor. The Manchester School of Economics and Social Studies, 22(2), 139–191. Lipton, M. (2004). Crop science, poverty and the family in a globalizing world. Plenary Session, Brisbane International Crop Science Conference. Maxwell, S., & Slater, R. (2003). Food policy old and new. Development Policy Review, 21(5), 531–553. Poulton, C., Kydd, J., & Dorward, A. (2006). Overcoming market constraints on pro-poor agricultural growth in sub-Saharan Africa. Development Policy Review, 24(3), 243–277. Reardon, T., Timmer, C. P., Barrett, C., & Berdeguè, J. (2003). The rise of supermarkets in Africa, Asia, and Latin America. American Journal of Agricultural Economics, 85(5), 1140–1146. 9 Rosegrant, M. W., Paisner, M. S., Meijer, S., & Witcover, J. (2001). Global food projections to 2020. International Food Policy Research Institute. Rosegrant, M. W., & Hazell, P. (2000). Transforming the rural Asian economy: The unfinished revolution. Hong Kong: Oxford University Press. Schiff, M., & Valdez, A. (1992). The plundering of agriculture in developing countries. Washington, DC: World Bank. Spencer, D. (1994). Infrastructure and technology constraints to agricultural development in the humid and subhumid tropics of Africa. Washington, DC: IFPRI. Thurlow, J., Kiringai, J., & Gautam, M. (2007). Rural investments to accelerate growth and poverty reduction in Kenya. Discussion paper no. 723. Washington, DC: International Food Policy Research Institute. Thurlow, J., & Wobst, P. (2006). Not all growth is equally good for the poor: The case of Zambia. Journal of African Economies, 15(4), 603–625. Timmer, C. P. (2005). Agriculture and pro-poor growth: What the literature says. Draft paper. World Bank, Washington, DC: Agricultural and Rural Development Department. Turner, B. L., II Hyden, G., & Kates, R. W. (Eds.) (1993). Population growth and agricultural change in Africa. Gainesville: University Press of Florida. Wiggins, S. (1995). Change in African farming systems between the mid1970s and the mid-1980s. Journal of International Development, 7(6), 807–848. Wiggins, S. (2000). Interpreting changes from the 1970s to the 1990s in African agriculture through village studies. World Development, 28(4), 631–662. World Bank (2006). World development indicators. Washington, DC: World Bank. Available online at www.sciencedirect.com Please cite this article in press as: Diao, X. et al. The Role of Agriculture in African Development, World Development (2010), doi:10.1016/j.worlddev.2009.06.011